Prior research indicates that testosterone is related both to antisocial behavior and to behavior that is neutral or even prosocial. The moderator variables that might lead testosterone to have either positive or negative effects have not been identified, and conditions that predict whether testosterone will be associated with prosocial or antisocial behavior remain unclear. There is also a more basic question of whether testosterone has any causal effects on behavior at all, or whether it simply varies in response to behavioral experiences. The proposed research will examine short-term effects of testosterone and consider how these might lead to different long-term effects in different individuals. The model to be tested regards testosterone as a source of transient arousal and focused attention that, depending upon an individual's interests and motives, lead toward actions that may be either prosocial or antisocial. Successful action that is gratifying to the individual will increase testosterone, and failed action that is discouraging will decrease testosterone, and these changes in turn will modify the impetus toward preferred behavior. In summary, testosterone will potentiate action that is congruent with an individual's personality and goals, and the consequences of this action will feed back to affect testosterone level, affecting further action. The model will be tested first with four experiments and a correlational study. The experiments will test the hypotheses that transdermal testosterone treatments will (a) increase boldness and feelings of dominance; (b) increase engagement with the environment and participation in social activity; (c) increase persistence and focused attention in performing laboratory tasks; and (d) increase favorability of implicit attitudes toward activity, competition, and aggression. The correlational study will examine dominance and nurturance as moderator variables that guide the effects of testosterone toward antisocial or prosocial behavior among college students.